“…Previous studies have prescribed surface forcing [e.g., Jumikis , 1977; Goodrich , 1978, 1982b; Lunardini , 1981; Riseborough , 2004], or prescribed soil moisture and thermal properties [e.g., Goodrich , 1978, 1982b; Romanovsky et al , 1997; Hinzman et al , 1998; Li and Koike , 2003; Ling and Zhang , 2004; Woo et al , 2004]. For example, in the validation of a modified Stefan's algorithm, a 5% unfrozen water content was used for six field sites ranged from polar desert to agricultural land [ Woo et al , 2004], while observed values ranged from 0% in coarse‐grained mineral soil to more than 20% in certain organic or clay soils [ Anderson and Tice , 1972; Nakano and Brown , 1972; Romanovsky and Osterkamp , 2000; Quinton et al , 2005; Overduin et al , 2006]. Most existing evaluation/validation studies of GTFD simulations only included single algorithm [e.g., Goodrich , 1978; Fox , 1992; Hinkel and Nicholas , 1995; Hinzman et al , 1998; Cherkauer and Lettenmaier , 1999; Quinton and Gray , 2001; Zhang et al , 2003; Woo et al , 2004; Hayashi et al , 2007], and only a few compared different algorithms [e.g., Romanovsky et al , 1997; Luo et al , 2003].…”